(subject to further revisions)
# | Date | Topic | Key concepts | Slides | Homework | Solutions | 4th hour projects |
1 | 01/20 | Introduction | Overview, Class policies | HW0 (not graded) | HW0 solutions | ||
2 | 01/22 | Supervised learning | Instance, label, and hypothesis spaces, learning algorithms | ||||
3 | 01/27 | Decision Trees | Entropy, Information Gain | ||||
4 | 01/29 | Linear classifiers | Linear separability, loss and risk, (stochastic) gradient descent | HW1 out | HW1 solutions | ||
5 | 02/03 | Bias/Variance tradeoff | Overfitting vs underfitting, bias and variance of a classifier | ||||
6 | 02/05 | Online learning algorithms | Perceptron and Winnow update rules | ||||
7 | 02/10 | Analyzing online learners | Concept learning, convergence and mistake bounds | ||||
8 | 02/12 | Kernels and Dual | Dual representation, Kernel trick | HW1 due; HW2 out | HW2 solutions | ||
9 | 02/17 | Hypothesis testing | t-Test etc. | ||||
10 | 02/19 | Max-margin classifiers | Tell us about your project. | ||||
11 | 02/24 | Support Vector Machines | |||||
12 | 02/26 | Multiclass classification | HW2 due; HW3 out | HW3 solutions | |||
13 | 03/03 | Midterm Review | Previous midterm, solutions | ||||
14 | 03/05 | Midterm Exam | |||||
15 | 03/10 | Multiclass classification | |||||
16 | 03/12 | Learning Theory | Concept learning, PAC learning, VC dimension | HW3 due; HW4 out | HW4 solutions | ||
17 | 03/17 | Ensemble Learning | Boosting and Bagging | ||||
18 | 03/19 | Probabilistic models | Graphical models, Bayes rule, MLE, MAP; Naive Bayes | Submit first report (template here) (Email to Julia) | |||
19 | 03/31 | Logistic regression | Tom Mitchell's notes | ||||
20 | 04/02 | Expectation-Maximization | EM algorithm, applied to mixture models, k-means | HW4 due; HW5 out | HW5 solutions | ||
21 | 04/07 | Hidden Markov Models | Viterbi | ||||
22 | 04/09 | Learning HMMs | Forward-Backward | ||||
23 | 04/14 | Still HMMs | |||||
24 | 04/16 |
|
Class canceled | HW5 due; HW6 out | HW6 solutions | ||
25 | 04/21 | The EM algorithm | Mixture models | ||||
26 | 04/23 | Clustering | k-means, agglomerative clustering, etc. | Submit second report | |||
27 | 04/28 | Wrap-up | |||||
28 | 04/30 | Final Review | HW6 due | ||||
29 | 05/05 | Final Exam | |||||
30 | 05/08 | Submit final report |